Datasets:

ArXiv:
DOI:
License:
zhuwq0 commited on
Commit
4023dde
2 Parent(s): 2c10c18 61b26a5

Merge branch 'main' of hf.co:datasets/AI4EPS/quakeflow_nc

Browse files
Files changed (3) hide show
  1. README.md +34 -60
  2. example.py +41 -0
  3. quakeflow_nc.py +28 -21
README.md CHANGED
@@ -66,7 +66,7 @@ Waveform data, metadata, or data products for this study were accessed through t
66
  - datasets
67
  - h5py
68
  - fsspec
69
- - torch (for PyTorch)
70
 
71
  ### Usage
72
  Import the necessary packages:
@@ -74,7 +74,6 @@ Import the necessary packages:
74
  import h5py
75
  import numpy as np
76
  import torch
77
- from torch.utils.data import Dataset, IterableDataset, DataLoader
78
  from datasets import load_dataset
79
  ```
80
  We have 6 configurations for the dataset:
@@ -89,16 +88,28 @@ We have 6 configurations for the dataset:
89
 
90
  The sample of `station` is a dictionary with the following keys:
91
  - `data`: the waveform with shape `(3, nt)`, the default time length is 8192
92
- - `phase_pick`: the probability of the phase pick with shape `(3, nt)`, the first dimension is noise, P and S
93
- - `event_location`: the event location with shape `(4,)`, including latitude, longitude, depth and time
 
 
 
 
 
 
 
94
  - `station_location`: the station location with shape `(3,)`, including latitude, longitude and depth
95
 
96
  The sample of `event` is a dictionary with the following keys:
97
  - `data`: the waveform with shape `(n_station, 3, nt)`, the default time length is 8192
98
- - `phase_pick`: the probability of the phase pick with shape `(n_station, 3, nt)`, the first dimension is noise, P and S
99
- - `event_center`: the probability of the event time with shape `(n_station, feature_nt)`, default feature time length is 512
100
- - `event_location`: the space-time coordinates of the event with shape `(n_staion, 4, feature_nt)`
101
- - `event_location_mask`: the probability mask of the event time with shape `(n_station, feature_nt)`
 
 
 
 
 
102
  - `station_location`: the space coordinates of the station with shape `(n_station, 3)`, including latitude, longitude and depth
103
 
104
  The default configuration is `station_test`. You can specify the configuration by argument `name`. For example:
@@ -117,70 +128,33 @@ quakeflow_nc = load_dataset("AI4EPS/quakeflow_nc", name="station_test", split="t
117
  quakeflow_nc = load_dataset("AI4EPS/quakeflow_nc", name="event", split="train")
118
  ```
119
 
120
- #### Usage for `station`
121
- Then you can change the dataset into PyTorch format iterable dataset, and view the first sample:
122
  ```python
123
  quakeflow_nc = load_dataset("AI4EPS/quakeflow_nc", name="station_test", split="test")
124
- # for PyTorch DataLoader, we need to divide the dataset into several shards
125
- num_workers=4
126
- quakeflow_nc = quakeflow_nc.to_iterable_dataset(num_shards=num_workers)
127
- # because add examples formatting to get tensors when using the "torch" format
128
- # has not been implemented yet, we need to manually add the formatting when using iterable dataset
129
- # if you want to use dataset directly, just use
130
- # quakeflow_nc.with_format("torch")
131
- quakeflow_nc = quakeflow_nc.map(lambda x: {key: torch.from_numpy(np.array(value, dtype=np.float32)) for key, value in x.items()})
132
- try:
133
- isinstance(quakeflow_nc, torch.utils.data.IterableDataset)
134
- except:
135
- raise Exception("quakeflow_nc is not an IterableDataset")
136
 
137
  # print the first sample of the iterable dataset
138
  for example in quakeflow_nc:
139
  print("\nIterable test\n")
140
  print(example.keys())
141
  for key in example.keys():
142
- print(key, example[key].shape, example[key].dtype)
 
 
 
143
  break
144
 
145
- dataloader = DataLoader(quakeflow_nc, batch_size=4, num_workers=num_workers)
 
 
146
 
147
  for batch in dataloader:
148
  print("\nDataloader test\n")
149
- print(batch.keys())
150
- for key in batch.keys():
151
- print(key, batch[key].shape, batch[key].dtype)
152
- break
153
- ```
154
-
155
- #### Usage for `event`
156
-
157
- Then you can change the dataset into PyTorch format dataset, and view the first sample (Don't forget to reorder the keys):
158
- ```python
159
- quakeflow_nc = datasets.load_dataset("AI4EPS/quakeflow_nc", split="test", name="event_test")
160
-
161
- # for PyTorch DataLoader, we need to divide the dataset into several shards
162
- num_workers=4
163
- quakeflow_nc = quakeflow_nc.to_iterable_dataset(num_shards=num_workers)
164
- quakeflow_nc = quakeflow_nc.map(lambda x: {key: torch.from_numpy(np.array(value, dtype=np.float32)) for key, value in x.items()})
165
- try:
166
- isinstance(quakeflow_nc, torch.utils.data.IterableDataset)
167
- except:
168
- raise Exception("quakeflow_nc is not an IterableDataset")
169
-
170
- # print the first sample of the iterable dataset
171
- for example in quakeflow_nc:
172
- print("\nIterable test\n")
173
- print(example.keys())
174
- for key in example.keys():
175
- print(key, example[key].shape, example[key].dtype)
176
- break
177
-
178
- dataloader = DataLoader(quakeflow_nc, batch_size=1, num_workers=num_workers)
179
-
180
- for batch in dataloader:
181
- print("\nDataloader test\n")
182
- print(batch.keys())
183
- for key in batch.keys():
184
- print(key, batch[key].shape, batch[key].dtype)
185
  break
186
  ```
 
66
  - datasets
67
  - h5py
68
  - fsspec
69
+ - pytorch
70
 
71
  ### Usage
72
  Import the necessary packages:
 
74
  import h5py
75
  import numpy as np
76
  import torch
 
77
  from datasets import load_dataset
78
  ```
79
  We have 6 configurations for the dataset:
 
88
 
89
  The sample of `station` is a dictionary with the following keys:
90
  - `data`: the waveform with shape `(3, nt)`, the default time length is 8192
91
+ - `begin_time`: the begin time of the waveform data
92
+ - `end_time`: the end time of the waveform data
93
+ - `phase_time`: the phase arrival time
94
+ - `phase_index`: the time point index of the phase arrival time
95
+ - `phase_type`: the phase type
96
+ - `phase_polarity`: the phase polarity in ('U', 'D', 'N')
97
+ - `event_time`: the event time
98
+ - `event_time_index`: the time point index of the event time
99
+ - `event_location`: the event location with shape `(3,)`, including latitude, longitude, depth
100
  - `station_location`: the station location with shape `(3,)`, including latitude, longitude and depth
101
 
102
  The sample of `event` is a dictionary with the following keys:
103
  - `data`: the waveform with shape `(n_station, 3, nt)`, the default time length is 8192
104
+ - `begin_time`: the begin time of the waveform data
105
+ - `end_time`: the end time of the waveform data
106
+ - `phase_time`: the phase arrival time with shape `(n_station,)`
107
+ - `phase_index`: the time point index of the phase arrival time with shape `(n_station,)`
108
+ - `phase_type`: the phase type with shape `(n_station,)`
109
+ - `phase_polarity`: the phase polarity in ('U', 'D', 'N') with shape `(n_station,)`
110
+ - `event_time`: the event time
111
+ - `event_time_index`: the time point index of the event time
112
+ - `event_location`: the space-time coordinates of the event with shape `(n_staion, 3)`
113
  - `station_location`: the space coordinates of the station with shape `(n_station, 3)`, including latitude, longitude and depth
114
 
115
  The default configuration is `station_test`. You can specify the configuration by argument `name`. For example:
 
128
  quakeflow_nc = load_dataset("AI4EPS/quakeflow_nc", name="event", split="train")
129
  ```
130
 
131
+ #### Example loading the dataset
 
132
  ```python
133
  quakeflow_nc = load_dataset("AI4EPS/quakeflow_nc", name="station_test", split="test")
 
 
 
 
 
 
 
 
 
 
 
 
134
 
135
  # print the first sample of the iterable dataset
136
  for example in quakeflow_nc:
137
  print("\nIterable test\n")
138
  print(example.keys())
139
  for key in example.keys():
140
+ if key == "data":
141
+ print(key, np.array(example[key]).shape)
142
+ else:
143
+ print(key, example[key])
144
  break
145
 
146
+ # %%
147
+ quakeflow_nc = quakeflow_nc.with_format("torch")
148
+ dataloader = DataLoader(quakeflow_nc, batch_size=8, num_workers=0, collate_fn=lambda x: x)
149
 
150
  for batch in dataloader:
151
  print("\nDataloader test\n")
152
+ print(f"Batch size: {len(batch)}")
153
+ print(batch[0].keys())
154
+ for key in batch[0].keys():
155
+ if key == "data":
156
+ print(key, np.array(batch[0][key]).shape)
157
+ else:
158
+ print(key, batch[0][key])
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
159
  break
160
  ```
example.py ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # %%
2
+ import numpy as np
3
+ from datasets import load_dataset
4
+ from torch.utils.data import DataLoader
5
+
6
+ # quakeflow_nc = load_dataset("AI4EPS/quakeflow_nc", name="station_test", split="test")
7
+ quakeflow_nc = load_dataset(
8
+ "./quakeflow_nc.py",
9
+ name="station_test",
10
+ # name="event_test",
11
+ split="test",
12
+ download_mode="force_redownload",
13
+ )
14
+
15
+ # print the first sample of the iterable dataset
16
+ for example in quakeflow_nc:
17
+ print("\nIterable test\n")
18
+ print(example.keys())
19
+ for key in example.keys():
20
+ if key == "data":
21
+ print(key, np.array(example[key]).shape)
22
+ else:
23
+ print(key, example[key])
24
+ break
25
+
26
+ # %%
27
+ quakeflow_nc = quakeflow_nc.with_format("torch")
28
+ dataloader = DataLoader(quakeflow_nc, batch_size=8, num_workers=0, collate_fn=lambda x: x)
29
+
30
+ for batch in dataloader:
31
+ print("\nDataloader test\n")
32
+ print(f"Batch size: {len(batch)}")
33
+ print(batch[0].keys())
34
+ for key in batch[0].keys():
35
+ if key == "data":
36
+ print(key, np.array(batch[0][key]).shape)
37
+ else:
38
+ print(key, batch[0][key])
39
+ break
40
+
41
+ # %%
quakeflow_nc.py CHANGED
@@ -50,7 +50,7 @@ _LICENSE = ""
50
  # TODO: Add link to the official dataset URLs here
51
  # The HuggingFace Datasets library doesn't host the datasets but only points to the original files.
52
  # This can be an arbitrary nested dict/list of URLs (see below in `_split_generators` method)
53
- _REPO = "https://huggingface.co/datasets/AI4EPS/quakeflow_nc/resolve/main/waveform_ps_h5"
54
  _FILES = [
55
  "1987.h5",
56
  "1988.h5",
@@ -104,14 +104,10 @@ class BatchBuilderConfig(datasets.BuilderConfig):
104
  """
105
  yield a batch of event-based sample, so the number of sample stations can vary among batches
106
  Batch Config for QuakeFlow_NC
107
- :param batch_size: number of samples in a batch
108
- :param num_stations_list: possible number of stations in a batch
109
  """
110
 
111
- def __init__(self, batch_size: int, num_stations_list: List, **kwargs):
112
  super().__init__(**kwargs)
113
- self.batch_size = batch_size
114
- self.num_stations_list = num_stations_list
115
 
116
 
117
  # TODO: Name of the dataset usually matches the script name with CamelCase instead of snake_case
@@ -172,13 +168,14 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
172
  features = datasets.Features(
173
  {
174
  "data": datasets.Array2D(shape=(3, self.nt), dtype="float32"),
175
- "phase_pick": datasets.Array2D(shape=(3, self.nt), dtype="float32"),
176
  "phase_time": datasets.Sequence(datasets.Value("string")),
177
  "phase_index": datasets.Sequence(datasets.Value("int32")),
178
  "phase_type": datasets.Sequence(datasets.Value("string")),
179
  "phase_polarity": datasets.Sequence(datasets.Value("string")),
180
  "begin_time": datasets.Value("string"),
181
  "end_time": datasets.Value("string"),
 
 
182
  "event_location": datasets.Sequence(datasets.Value("float32")),
183
  "station_location": datasets.Sequence(datasets.Value("float32")),
184
  },
@@ -186,18 +183,21 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
186
  elif (self.config.name == "event") or (self.config.name == "event_train") or (self.config.name == "event_test"):
187
  features = datasets.Features(
188
  {
189
- "data": datasets.Array2D(shape=(3, self.nt), dtype="float32"),
190
- "phase_pick": datasets.Array2D(shape=(3, self.nt), dtype="float32"),
191
- "phase_time": datasets.Sequence(datasets.Value("string")),
192
- "phase_index": datasets.Sequence(datasets.Value("int32")),
193
- "phase_type": datasets.Sequence(datasets.Value("string")),
194
- "phase_polarity": datasets.Sequence(datasets.Value("string")),
195
  "begin_time": datasets.Value("string"),
196
  "end_time": datasets.Value("string"),
 
 
197
  "event_location": datasets.Sequence(datasets.Value("float32")),
198
- "station_location": datasets.Sequence(datasets.Value("float32")),
199
  },
200
  )
 
 
201
 
202
  return datasets.DatasetInfo(
203
  # This is the description that will appear on the datasets page.
@@ -224,7 +224,8 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
224
  # By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
225
  urls = _URLS[self.config.name]
226
  # files = dl_manager.download(urls)
227
- files = dl_manager.download_and_extract(urls)
 
228
  print(files)
229
 
230
  if self.config.name == "station" or self.config.name == "event":
@@ -280,9 +281,12 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
280
  event_attrs["longitude"],
281
  event_attrs["latitude"],
282
  event_attrs["depth_km"],
283
- event_attrs["event_time_index"],
284
  ]
 
 
285
  station_ids = list(event.keys())
 
 
286
  if (
287
  (self.config.name == "station")
288
  or (self.config.name == "station_train")
@@ -307,6 +311,8 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
307
  "phase_polarity": phase_polarity,
308
  "begin_time": begin_time,
309
  "end_time": end_time,
 
 
310
  "event_location": event_location,
311
  "station_location": station_location,
312
  }
@@ -327,14 +333,13 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
327
  for i, sta_id in enumerate(station_ids):
328
  waveforms[i, :, : self.nt] = event[sta_id][:, : self.nt]
329
  attrs = event[sta_id].attrs
330
- phase_type.append(attrs["phase_type"])
331
- phase_time.append(attrs["phase_time"])
332
- phase_index.append(attrs["phase_index"])
333
- phase_polarity.append(attrs["phase_polarity"])
334
  station_location.append(
335
  [attrs["longitude"], attrs["latitude"], -attrs["elevation_m"] / 1e3]
336
  )
337
-
338
  yield event_id, {
339
  "data": waveforms,
340
  "phase_time": phase_time,
@@ -343,6 +348,8 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
343
  "phase_polarity": phase_polarity,
344
  "begin_time": begin_time,
345
  "end_time": end_time,
 
 
346
  "event_location": event_location,
347
  "station_location": station_location,
348
  }
 
50
  # TODO: Add link to the official dataset URLs here
51
  # The HuggingFace Datasets library doesn't host the datasets but only points to the original files.
52
  # This can be an arbitrary nested dict/list of URLs (see below in `_split_generators` method)
53
+ _REPO = "https://huggingface.co/datasets/AI4EPS/quakeflow_nc/resolve/main/waveform_h5"
54
  _FILES = [
55
  "1987.h5",
56
  "1988.h5",
 
104
  """
105
  yield a batch of event-based sample, so the number of sample stations can vary among batches
106
  Batch Config for QuakeFlow_NC
 
 
107
  """
108
 
109
+ def __init__(self, **kwargs):
110
  super().__init__(**kwargs)
 
 
111
 
112
 
113
  # TODO: Name of the dataset usually matches the script name with CamelCase instead of snake_case
 
168
  features = datasets.Features(
169
  {
170
  "data": datasets.Array2D(shape=(3, self.nt), dtype="float32"),
 
171
  "phase_time": datasets.Sequence(datasets.Value("string")),
172
  "phase_index": datasets.Sequence(datasets.Value("int32")),
173
  "phase_type": datasets.Sequence(datasets.Value("string")),
174
  "phase_polarity": datasets.Sequence(datasets.Value("string")),
175
  "begin_time": datasets.Value("string"),
176
  "end_time": datasets.Value("string"),
177
+ "event_time": datasets.Value("string"),
178
+ "event_time_index": datasets.Value("int32"),
179
  "event_location": datasets.Sequence(datasets.Value("float32")),
180
  "station_location": datasets.Sequence(datasets.Value("float32")),
181
  },
 
183
  elif (self.config.name == "event") or (self.config.name == "event_train") or (self.config.name == "event_test"):
184
  features = datasets.Features(
185
  {
186
+ "data": datasets.Array3D(shape=(None, 3, self.nt), dtype="float32"),
187
+ "phase_time": datasets.Sequence(datasets.Sequence(datasets.Value("string"))),
188
+ "phase_index": datasets.Sequence(datasets.Sequence(datasets.Value("int32"))),
189
+ "phase_type": datasets.Sequence(datasets.Sequence(datasets.Value("string"))),
190
+ "phase_polarity": datasets.Sequence(datasets.Sequence(datasets.Value("string"))),
 
191
  "begin_time": datasets.Value("string"),
192
  "end_time": datasets.Value("string"),
193
+ "event_time": datasets.Value("string"),
194
+ "event_time_index": datasets.Value("int32"),
195
  "event_location": datasets.Sequence(datasets.Value("float32")),
196
+ "station_location": datasets.Sequence(datasets.Sequence(datasets.Value("float32"))),
197
  },
198
  )
199
+ else:
200
+ raise ValueError(f"config.name = {self.config.name} is not in BUILDER_CONFIGS")
201
 
202
  return datasets.DatasetInfo(
203
  # This is the description that will appear on the datasets page.
 
224
  # By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
225
  urls = _URLS[self.config.name]
226
  # files = dl_manager.download(urls)
227
+ # files = dl_manager.download_and_extract(urls)
228
+ files = ["waveform_h5/1989.h5", "waveform_h5/1990.h5"]
229
  print(files)
230
 
231
  if self.config.name == "station" or self.config.name == "event":
 
281
  event_attrs["longitude"],
282
  event_attrs["latitude"],
283
  event_attrs["depth_km"],
 
284
  ]
285
+ event_time = event_attrs["event_time"]
286
+ event_time_index = event_attrs["event_time_index"]
287
  station_ids = list(event.keys())
288
+ if len(station_ids) == 0:
289
+ continue
290
  if (
291
  (self.config.name == "station")
292
  or (self.config.name == "station_train")
 
311
  "phase_polarity": phase_polarity,
312
  "begin_time": begin_time,
313
  "end_time": end_time,
314
+ "event_time": event_time,
315
+ "event_time_index": event_time_index,
316
  "event_location": event_location,
317
  "station_location": station_location,
318
  }
 
333
  for i, sta_id in enumerate(station_ids):
334
  waveforms[i, :, : self.nt] = event[sta_id][:, : self.nt]
335
  attrs = event[sta_id].attrs
336
+ phase_type.append(list(attrs["phase_type"]))
337
+ phase_time.append(list(attrs["phase_time"]))
338
+ phase_index.append(list(attrs["phase_index"]))
339
+ phase_polarity.append(list(attrs["phase_polarity"]))
340
  station_location.append(
341
  [attrs["longitude"], attrs["latitude"], -attrs["elevation_m"] / 1e3]
342
  )
 
343
  yield event_id, {
344
  "data": waveforms,
345
  "phase_time": phase_time,
 
348
  "phase_polarity": phase_polarity,
349
  "begin_time": begin_time,
350
  "end_time": end_time,
351
+ "event_time": event_time,
352
+ "event_time_index": event_time_index,
353
  "event_location": event_location,
354
  "station_location": station_location,
355
  }